Autocrine regulation of cell proliferation by estrogen receptor-alpha in estrogen receptor-alpha-positive breast cancer cell lines

Antibodies for ERα (F10), Ki-67 (H-300), cyclin A (C-19), Cyclin B (H-20), cyclin D1 (H-295), cyclin E (C-19) were obtained from Santa Cruz (Santa Cruz, CA). Blocking peptides for ERα (F10), cyclin A (C-19), cyclin E (C-19) were obtained from Santa Cruz (Santa Cruz, CA). Antibodies for ERK1/2 (137F5), phospho-ERK1/2 (E10), AKT, phospho-AKT/Ser473, and EGFR were obtained from Cell Signaling Technology (Danvers, MA). Secondary antibody conjugated with Alexa Fluor 488 (Molecular Probes, Inc, Eugene, OR) was used for green fluorescent staining, secondary antibody conjugated with Rhodamine Red-X (Jackson ImmunoResearch, West Grove, PA) was used for red fluorescent staining. Heat inactivated FBS and Charcoal-Dextrin stripped FBS (CD-FBS) were obtained from Omega (Tarzana, CA). Estrogen was obtained from Sigma (St. Louis, MO). ICI182780 was obtained from TOCRIS (Ellisville, MO). Gefitinib was obtained from LC Laboratories (Woburn, MA). EGF was obtained from Invitrogen (Carlsbad, CA).

DMEM/F12 medium (MediaTech Inc, Herndon, VA) containing 10% heat inactivated FBS and 5 μg/ml insulin was used for routine maintenance of the three ERα-positive breast cancer cell lines, MCF-7, T47D, and ZR75-1 (ATCC). For experimental assays, phenol-red free DMEM/F12 (MediaTech Inc, Herndon, VA) containing 5% CD-FBS was used unless otherwise specified. To arrest cells in G1 phase by ICI182780, cells were seeded in culture dishes for 1–2 days to let the cells reach about 30–40% confluency before replacement with medium containing ICI182780. To evaluate the effect of estrogen stimulation, cells were stimulated by adding E2 either after removal of ICI182780 or in the presence of ICI182780. For estrogen stimulation after removal of ICI182780, cells treated with ICI182780 were washed briefly two times with serum-free phenol red-free DMEM/F12 to remove ICI182780, followed by replacement with medium containing 5 nM estrogen but without ICI182780. The control cells were replaced with medium containing vehicle only. For estrogen stimulation in the presence of ICI182780, cells were treated with 10 nM ICI182780 for various times before adding ten fold of E2 (100 nM) to the cells without removing ICI182780. ICI182780 stock solution was made at 10 mM in DMSO, then serially diluted in DMSO to 10 μM before diluted in culture medium. Estrogen stock solution was made at 10 mM in ethanol, then serially diluted in ethanol to 10 μM before diluted in culture medium. To inhibit the EGFR activation, cells were treated with Gefitinib for 2 hr first; without removing Gefitinib, estrogen or EGF was then added to the cells for stimulation. Gefitinib stock was made in DMSO at 1–50 mM, the stock was diluted in culture medium to make the final concentrations at 1–50 μM.

Cells grown on glass coverslips were briefly washed with ice-cold Dulbecco's PBS and fixed with cold 4% paraformaldehyde containing 0.01% Triton X-100 for 25 min on ice. After washing once briefly with PBS-T (Dulbelcco's PBS containing 0.01% Triton X-100), the cells were treated with 0.05% Triton x-100 in PBS for 10 min for permeabilization. The cells were washed once briefly with PBS-T, followed by blocking in PBS-T containing 5% normal serum and 50 mM NH4Cl for 30 min at room temperature. The cells were then incubated with primary antibodies diluted in PBS-T containing 1% normal serum. Anti-ERα antibody was diluted at 1:50, anti-Ki-67 at 1:100, anti-cyclin D at 1:150, anti-cyclin E at 1:100, anti-cyclin A at 1:100, and anti-cyclin B at 1:100. For the antibodies with blocking peptides available, the antibodies (ERα, cyclin A, cyclin E) were evaluated for its binding specificity by incubation with 5 fold of blocking peptides before applying to cells for primary antibody incubation. After overnight incubation at 4°C, the cells were washed four times with PBS-T, 7 min each time. Then the cells were incubated for 1 hr at room temperature with fluorescent dye-conjugated secondary antibodies (1:200) in PBS-T containing 1% normal serum. The cells were washed 4 times with PBS-T, 7 min each time before mounted on slides with VectoShield with DAPI (Vector Laboratories, Burlingame, CA). Immunofluorescent staining(s) was observed under Olympus BX50 Fluorescence Microscope with Optronics MagnaFire digital camera (Microscope Image Center, UVM).

For analysis of cell cycle distribution, cells were harvested by trypsinization, pelleted by centrifugation, and washed once with Dulbecco's PBS. While vortexing cells, ice-cold 70–75% ethanol was added drop wise to resuspend the cells. The cells were fixed in 70% ethanol for at least 1 hr at 4°C before staining or stored at -20°C freezer. Before staining, 1–3 millions of cells were pelleted, washed once with PBS, followed by resuspension in 1 ml PBS containing 20 μg/ml PI (propidium iodide), 150 μg/ml RNaseA, 10 mM EDTA, and 0.1% Triton X-100. Incubation was carried in the dark at 37°C for 15–30 min. For the analysis of cell cycle distribution of ERα-staining cells, cells were harvested by trypsinization and the cell number counted. The cells were pelleted by centrifugation (2000 rpm, 5 min, 4°C). After resuspension with culture medium, 3 million of cells were transferred into Ependorff tube and pelleted by centrifugation (500 × g, 4°C, 2 min). The cell pellet was resuspended in 1 ml PBS-T containing 1% paraformaldehyde and cells were fixed on ice for 10 min. The cells were pelleted by brief centrifugation, resuspended in PBS containing 0.05% Triton-X100 for permeabilization at room temperature for 5 min. After brief washing, cells were incubated with 5% normal serum in PBS-T (30 min, room temperature) for blocking. Cells were pelleted and resuspended in PBS-T containing anti-ERα (F10, 1:100 dilution) and 0.5% BSA, followed by incubation at 4°C for overnight. The cells were washed 4 times with PBS-T, 7 min each, by centrifugation and resuspension, followed by incubation with the secondary antibody anti-mouse Alexa Fluor 488 (1:300) at room temperature for 40 min. After washing with PBS-T for 4 times, 7 min each, the cells were incubated with 0.5 ml PI-RNaseA solution for 30 min at 37°C. Flow cytometry analysis was performed using the Coulter Epics XL-MCL (VCC core facility, UVM). About 20,000 cells were analyzed for each sample.

Cells at 70–80% confluence were washed briefly twice with ice-cold D-PBS before scraping on ice with lysis buffer (20 mM Tris-HCl, pH 7.5; 150 mM NaCl; 2.5 mM sodium pyrophosphate; 1 mM sodium β-glycerophosphate; 5 mM NaF, 1 mM Na3VO4, 1 mM phenylmethylsulfonyl fluoride, 1% Triton X-100, and 1 tablet of protease inhibitor mixture (Roche Molecular Biochemicals, Indianapolis, IN) per 40 ml of lysis buffer). Cellular debris was removed by centrifugation (14,000 × g for 15 min at 4°C). For SDS-PAGE, total cell lysate containing ~30 μg proteins were separated on 12% SDS-PAGE gel and transferred onto PVDF membranes (Millipore, Bedford, MA). Immunoblotting was carried out followed standard procedures [37, 38].

ERα is not colocalized with the cell proliferation marker Ki-67 in normal mammary cells, and colocalization of ERα and Ki-67 is rare even in primary breast tumors [1‐3, 16‐18, 23]. To study the mechanism of how cell proliferation is mediated by ERα, the ERα-positive breast cancer cell lines were used in our study. In all the three ERα-positive cell lines evaluated in this study, i.e., MCF-7, T47D, and ZR75-1, a large proportion of cells showed duel-staining of ERα and Ki-67 (Fig. 1). The colocalization of ERα with Ki-67 was evaluated in cells under different culture conditions and the duel-staining was observed in all culture conditions, including cell culture in regular DMEM/F12 supplemented with 10% FBS, phenol-red free DMEM/F12 supplemented with charcoal dextrin stripped FBS, with and without insulin, with and without estrogen. While these data don't exclude the possibility of a paracrine regulation of cell proliferation by ERα in these cells, the colocalization of ERα and Ki-67 does support the hypothesis that ERα may mediate cell proliferation in an autocrine mode in these ERα-positive breast cancer cell lines, a mechanism absent in normal mammary cells and most primary breast tumors [1‐3, 16‐18, 23, 24].

In normal breast tissue, ERα is expressed in quiescent (G₀) cells or early G₁ phase [1, 17]. In MCF-7 cells, activation of ERα leads to cyclin D1 and myc expression, indicating the expression of ERα in G1 phase [26]. Ki-67 is a proliferation marker expressed in late G1, S, G2, and M phases of the cell cycle [39]. In these ERα-positive breast cancer cell lines, most Ki-67-staining cells contained high levels of ERα suggesting that ERα might be present throughout the cell cycle progression in these cell lines (Fig. 1). To address this possibility, we evaluated the colocalization of ERα with the cyclins for different phases of cell cycle (cyclin D for G1 phase; cyclin E for late G1 phase; cyclin A for G1/S/G2 phases; cyclin B for M phase) [40‐42]. The cyclins for the different phases of cell cycle were found colocalized with a proportion of ERα cells, with the cyclin A showing the largest and cyclin B the least proportion of ERα-staining cells (Fig. 2A). Furthermore, flow cytometry analysis demonstrated that MCF-7 cells with high ERα expression (ERα^high) were distributed across G0/G1, S, and G2/M phases (Fig. 2B). These data indicate that ERα expression during cell cycle progression is deranged in MCF-7 cells.

While the distribution of ERα across all phases of cell cycling supports the autocrine mode of action in cell proliferation, it also raises the possibility that the colocalization of ERα with Ki-67 might be a coincidence. To further determine whether ERα can mediate cell proliferation in an autocrine mode, we sought to arrest the cells with the pure ERα antagonist ICI182780 followed by estrogen stimulation. When treated with ICI182780 for 72–96 hrs, no cell was observed with high ERα expression and only a few cells were stained with Ki-67, indicating that both ERα and Ki-67 were down-regulated by ICI182780 (Fig. 3). Consistently, ICI182780 treatment leads to cell cycle arrest at G1 phase (Fig. 3). The inhibitory effect of ICI182780 was observed similarly at concentrations from 10 nM to 50 nM, therefore 10 nM was selected for our later experiments.

To evaluate the effect of estrogen stimulation, cells were stimulated by adding E2 either after removal of ICI182780 or in the presence of ICI182780. We first evaluated the effect of estrogen stimulation after removal of ICI182780. MCF-7 cells were treated for ICI182780 for 72–96 hr and the treatment was stopped by washing with fresh serum-free medium to remove ICI182780. The cells were then stimulated with estrogen by adding medium containing 5 nM estrogen but without ICI182780. The control group of cells was replaced with medium without estrogen. In the cells stimulated with estrogen, both ERα and Ki-67 were induced with ERα and Ki-67 colocalized in most cells (Fig. 4A). In the control group of cells without estrogen stimulation, no high level of ERα or Ki-67 was observed although the cells were released from ICI182780 treatment. We also determined the effect of estrogen stimulation by adding estrogen directly to the cells still under ICI182780 treatment (Fig. 4B). The continuous presence of ICI182780 was to help minimize the secondary factors that might promote cell proliferation. Cells were treated with 10 nM ICI182780 for 72–96 hrs before the estrogen was added to the cells for a final concentration of 100 nM. Similar to the result shown in Fig. 4A, excessive estrogen stimulation in the presence of ICI182780 also induced the expression of high levels of ERα and Ki-67 with the two proteins colocalized in most cells (Fig. 4B). Collectively, these data indicate that activation of ERα by estrogen in ICI182780 arrested cells induces the expression of high levels of ERα and Ki-67 and their colocalization, supporting the autocrine mode of ERα action.

We next examined the effect of estrogen stimulation on cell cycle progression. As described above, cells were stimulated with estrogen either after release from ICI182780, or estrogen was added at 10 fold higher concentration (100 nM E2) directly to cells without removing ICI182780 (10 nM ICI). Similar results were observed for either stimulation (Fig. 5, and data not shown). In cells stimulated with estrogen after release from ICI182780, most cells were still in the G1 phase until 24 hr after stimulation. At 44 hr after estrogen stimulation, about 45% of the cells were in S or G2/M phases, indicating that activation of ERα promotes the progressing of the cell cycle (Fig. 5). Interestingly, induction of high levels of ERα and Ki-67 appeared much earlier, with high levels of ERα appeared at around 6 hr. It is noteworthy that high levels of ERα appeared earlier than Ki-67, supporting that the activation of ERα is mediating the expression of Ki-67 and the cell cycle progression. These data indicate that the colocalization of ERα and Ki-67 is not a coincidence and support the hypothesis that activation of ERα can promote cell proliferation by autocrine regulation.

In the paracrine regulation of cell proliferation by ERα, EGFR signaling is required [20]. Therefore we evaluated whether EGFR signaling is required for the autocrine regulation of cell proliferation by ERα in MCF-7 cells. EGFR inhibitor Gefitinib was used to block EGFR signaling, 20 μM of Gefitinib was chosen based on our preliminary experiments using a wide range concentrations of Gefitinib (1 μM – 50 μM) (Fig. 6D, and data not shown). We first evaluated whether Gefitinib treatment inhibits E2 stimulation induced expression and colocalization of ERα and Ki-67. MCF-7 cells were treated first with 10 nM ICI182780 for 72–96 hr, followed by adding the EGFR inhibitor Gefitinib to a final concentration of 20 μM for 2 hr, then by adding estrogen to a final concentration of 100 nM. In the presence of ICI182780 and Gefitinib, estrogen stimulation still induced the expression and colocalization of ERα and Ki-67 (Fig. 6B). Conceivably, Gefitinib treatment does not inhibit estrogen stimulation induced cell cycle progression (Fig. 6C lower panel). In the control groups of MCF-7 cells treated with ICI182780 for 72–96 hr, ERα and Ki-67 were down-regulated and the cells remained at G1 phase (Fig. 6A, C upper panel). In the growth curve studies and the EGF stimulation experiments, Gefitinib at 20 μM is sufficient to block EGFR signaling but not toxic to the cells (Fig. 6D, and data not shown). From these data, we concluded that EGFR signaling is not required for the autocrine action of ERα in mediating cell proliferation in MCF-7 cells.